Multivariate Analyses Reveal Biological Components Related to Neuronal Signaling and Immunity Mediating Electroencephalograms Abnormalities in Alcohol‐Dependent Individuals from the Collaborative Study on the Genetics of Alcoholism Cohort. (21st May 2019)

• Record Type: Journal Article
• Authors: Meda, Shashwath A.; Narayanan, Balaji; Chorlian, David; Meyers, Jacquelyn L.; Gelernter, Joel; Hesselbrock, Victor; Bauer, Lance; Calhoun, Vince D.; Porjesz, Bernice; Pearlson, Godfrey D.
• Is Part Of: Alcoholism. Volume 43:Number 7(2019)
• Page Start: 1462
• Page End: 1477
• Publication Date: 2019-05-21
• Subjects: Alcoholism -- Substance Use -- Parallel ICA -- Function -- Brain -- Collaborative Study on the Genetics of Alcoholism; Alcoholism -- Periodicals; Alcoholism -- Periodicals; Alcoolisme; Electronic journals; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.861005
• DOI: 10.1111/acer.14063
• Languages: English
• ISSNs: 0145-6008

Abstract

Abstract : Background: The underlying molecular mechanisms associated with alcohol use disorder (AUD) risk have only been partially revealed using traditional approaches such as univariate genomewide association and linkage‐based analyses. We therefore aimed to identify gene clusters related to Electroencephalograms (EEG) neurobiological phenotypes distinctive to individuals with AUD using a multivariate approach. Methods: The current project adopted a bimultivariate data‐driven approach, parallel independent component analysis (para‐ICA), to derive and explore significant genotype–phenotype associations in a case–control subset of the Collaborative Study on the Genetics of Alcoholism (COGA) dataset. Para‐ICA subjects comprised N = 799 self‐reported European Americans (367 controls and 432 AUD cases), recruited from COGA, who had undergone resting EEG and genotyping. Both EEG and genomewide single nucleotide polymorphism (SNP) data were preprocessed prior to being subjected to para‐ICA in order to derive genotype–phenotype relationships. Results: From the data, 4 EEG frequency and 4 SNP components were estimated, with 2 significantly correlated EEG–genetic relationship pairs. The first such pair primarily represented theta activity, negatively correlated with a genetic cluster enriched for (but not limited to) ontologies/disease processes representing cell signaling, neurogenesis, transmembrane drug transportation, alcoholism, and lipid/cholesterol metabolism. The second component pair represented mainly alpha activity, positively correlated with a genetic cluster with ontologies similarly enriched as the first component. Disease‐related enrichments for this component revealed heart and autoimmune disorders as top hits. Loading coefficients for both the alpha and theta components were significantly reduced in cases compared to controls. Conclusions: Our data suggest plausible multifactorial genetic components, primarily enriched for neuronal/synaptic signaling/transmission, immunity, and neurogenesis, mediating low‐frequency alpha and theta abnormalities in alcohol addiction.